Vitamin and mineral supplements are among the most commonly used drugs in the U.S. Neither chemoprevention trials nor observational studies have established the benefits or risks of supplemental use. The aim of this proposal is to investigate the association of intake of supplemental vitamin C, vitamin E, calcium and multivitamins with total cancer incidence. To meet this aim, a cohort will be recruited within the 13 counties of western Washington State, and followed for a mean of 2 1/4 years. A mailing will be sent to 300,000 men and women age 50-74 using names obtained from a commercial mailing list, which will include a recruitment letter targeting supplemental users and a questionnaire. The recruitment procedures have been tested, and should yield 35,000 male respondents and 40,000 female respondents with 75% of respondents having used at least one dietary supplement in the prior 10 years. The questionnaire solicits detailed information on supplement use over the last 10 years, and information on covariates that may be associated with supplement use and with future risk of cancer (e.g., medical history, cancer risk factors, cancer screening, reasons for supplement use). The questionnaire also incorporates a food frequency questionnaire, with additional items on supplemented foods. A second mailing to respondents 2-4 months later will supply a tape measure for anthropometric measures and brushes for collection of DNA from cheek cells. Studies to quantify the relative validity of our primary exposure measures will be conducted. Endpoint information will be efficiently and accurately collected by linking participant identifiers to the system Washington SEER cancer registry and the Washington State death tapes. Out-migration from the attachment areas of these files will be monitored by linkage to the National Change of Address tape. Longer term follow-up of this cohort will allow the investigation of the association of supplements with specific cancers (lung, prostate, breast and large bowel) and with total mortality, and will allow investigation of gene-supplement interactions. If supplements are harmful or of no use, this information would be important for the large number of Americans taking supplements. If a beneficial effect is found, it could be translated into highly cost effective cancer control measures.